Dyablo: models for the dynamic behaviour of Biological Networks

Principal Investigator of the IST team: Isabel Sá-Correia

Contract: PTDC/EIA/71587/2006

Start date: 01.01.2008

Duration: 36 months 


Much of the existing knowledge in molecular biology comes from detailed studies of specific componentes of biological systems. In contrast to much of molecular biology, systems biology does not seek to break down a system into its parts and study one part of the process at a time. Using knowledge from molecular biology, systems biology proposes hypotheses that explain the behavior of the complete system. Importantly, these hypotheses can be used to mathematically model the system and make predictions on how different changes in the environment affect the system. These predictions can be iteratively tested for their validity. Systems biology requires the development of new approaches by mathematicians, computer scientists, engineers, biologists and physicists, to improve our ability to make these high-throughput measurements and create, refine, and retest the models until the predicted behavior accurately reflects the experimentally observed reality. Complex biological systems can be analyzed as biochemical networks. These networks may represent protein-protein interactions, the metabolism of an organism, its system of gene expression regulation, or even, mixed networks that contain information coming from various of the previous sources. In the context of this project, we will concentrate our attention on the view of a cell or organism as a system of interconnected biochemical and genetic regulatory networks. Genetic, metabolic and protein-protein networks can be seen as networks of interactions between objects (genes, proteins and metabolites) whose deeper nature is not important in a first approximation. This project will be centered of the development of models for the dynamic behavior of this type of networks. These models will require detailed information about the structure and the parameters of the networks. Altough inferring the structure of the networks in a reliable way remains much of an open problem, we will assume that, for the cases under study, the structure is know, at least within some approximation, and that one is interested in modeling, with sufficient accuracy, the dynamic of the networks. The most accurate modeling of networks of this type requires the use of differential equations that model, with high accuracy, the physical phenomena that underlie the network dynamics. Regretably, this level of detail requires the identification of a high number of parameters, and is very limited in the size and complexity of the networks that can be tackled. A large number of results has shown that more abstract models can be used to make accurate predictions about the behavior of complex systems, since these systems seem to be robust against changes in the particular concentration values of biochemical components. In this project we will study the development and application of higher level models to complex biochemical networks, that will range in complexity from networks with less than a dozen components, to complete genetic regulatory networks in moderately complex organisms. This work will be anchored in previous results obtained by the research team on the development of models for biological networks. This work includes simulators for the qualitative analysis of genetic regulatory networks using piecewise afine differential equations models [de Jong et al. 2002] and its discrete time counterparts [Caldas et al, 2006]. These formalisms are useful when the target is a complex system, with possible complex dynamic, that, however, can be abstracted and studied in a qualitative way in order to obtain the characteristics of its state space. This project will develop methods for modeling the dynamics of biochemical networks and new approaches for the analysis of the characteristics of the state spaces of complex biological systems that will enable biologists and other scientists to characterize their properties. This will make it possible not only to model behaviors observed in vitro, but also to predict the behaviors of the systems when the experimental conditions are changed, either by modifying the environment, or by modifying the charateristics of the systems themselves. The availability of sufficiently accurate models will make it possible for biologists to conduct what-if experiments in large scale. The ability to perform this type of in-silico experiments will advance enourmously our capacity to study complex biological systems. This will have a strong impact in our ability to understand the behavior of complex organisms and to tackle one of the central challenges of biotechnology in the XXI century, that of synthetic biology.